Evaluation of Plantar Pressure Sensors for Classification of Ski Gear Using Deep Learning Models
摘要
This work focuses on the classification of different types of cross-country ski gear using instrumented insoles with a minimal number of pressure sensors placed inside ski boots. Two configurations were evaluated, using two or three sensors per insole. A deep learning model was used, which demonstrated promising results with two skiers in real-world evaluations, obtaining a minimum accuracy of 86% with three pressure sensors and 70% with two sensors per insole. Therefore, this work consists of an evaluation of the ability of machine learning and these wearables to classify gears in cross-country skiing in the skating style.